arXiv Research Unveils In-Context Learning Mechanisms: From Copying to Compositional Behavior in Language Models
A new arXiv study investigates the mechanisms of in-context learning (ICL) in language mod…
ZICQ curates daily AI news and deep dives on models, open-source infrastructure, and developer field reports.
Submit an articleCurrent filter: Tag: #语言识别 Clear filters
A new arXiv study investigates the mechanisms of in-context learning (ICL) in language mod…
Research & Papers· 2026-10-15
arXiv has released a study on Protein Class-Incremental Learning (CIL), systematically eva…
LLMs & Foundation Models· 2026-10-15
A new study published on arXiv introduces a self-distillation technique using inverse reas…
LLMs & Foundation Models· 2026-10-15
A new arXiv study examines the limitations of the Bradley-Terry model in Chatbot Arena, wh…
LLMs & Foundation Models· 2026-10-15
arXiv has released a study on the application of pre-trained language models in social sci…
Research & Papers· 2026-10-15
arXiv has released a study analyzing the linguistic representation of gender in science fi…
LLMs & Foundation Models· 2026-10-15
arXiv introduces Joint Subspace Sparse Attention (JSSA), a novel method addressing the KV-…
LLMs & Foundation Models· 2026-10-15
arXiv has released a new study on the event prediction capabilities of large language mode…
LLMs & Foundation Models· 2026-10-15
arXiv has released a study evaluating eight closed-source and open-weight large language m…
LLMs & Foundation Models· 2026-10-15
arXiv has released a study on CTI-REALM, examining the performance and cost trade-offs bet…
Agentic· 2026-10-15
The Qwen team has introduced ClawLoop, a novel framework for self-improving AI agents that…
Agentic· 2026-10-15
A new arXiv paper provides an in-depth analysis of Agentic Data Science (agentic DS), eval…